T Search needs about 16.4 GB of memory at its main file (a 15.2 GB file) with an 8K context window, so it fits on a 24 GB GPU. Longer context needs more: at 32K tokens the KV cache alone adds about 2.8 GB.
T Search – 35.5B agentic retrieval model for multi‑step search
T Search is a 35.5B parameter mixture‑of‑experts model built on Qwen3.6‑35B‑A3B. It performs agentic retrieval, issuing queries, inspecting results, and returning ranked evidence sets. The model supports reasoning and vision tasks and is trained for multi‑round search in English and Russian.
Good at
- Agentic multi‑round search
- Evidence ranking
- Reasoning with vision
Limits
- Requires 16.4 GB VRAM at 8 k context
- No vision support beyond bundled projector
- License is apache‑2.0
Best for
- Building search‑based agents
- Retrieval‑augmented generation pipelines
- Multi‑step evidence synthesis
Who it’s for. Developers building retrieval agents or search‑augmented LLM applications. The model is suitable for users who need robust multi‑round search and evidence ranking in English and Russian.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 15.2 GB | 16.4 GB | Too big | Get file → |
The only available GGUF file is the unknown quant version, 15.23 GB in size. It requires approximately 16.4 GB VRAM at 8 k context. For lower memory budgets, no other quantizations are provided.
What GPU do you need for T Search?
These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: no quant fits at an 8K context.
- 6 GB VRAM: no quant fits at an 8K context.
- 8 GB VRAM: no quant fits at an 8K context.
- 12 GB VRAM: no quant fits at an 8K context.
- 16 GB VRAM: no quant fits at an 8K context.
- 24 GB VRAM: the best quant that fits is the main file (15.2 GB file, about 16.4 GB in use).
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about T Search
How much VRAM does T Search need?
About 16.4 GB at its main file with an 8K context window. That is the 15.2 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run T Search on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 16.4 GB, so you would need a larger GPU or CPU offloading.
What is the file size of T Search GGUF?
It depends on the quant: file is 15.2 GB.
What context length does T Search support?
The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 0.7 GB at 8K and 2.8 GB at 32K.
Where can I download T Search GGUF?
The files are on Hugging Face in the IsValorum/T-Search-APEX-I-MiniPlus-V2.1-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does T Search use?
Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run T Search on 8 GB VRAM?
No. The model requires about 16.4 GB VRAM at 8 k context, so 8 GB is insufficient.
Does T Search support vision tasks?
Yes, it includes a bundled Q8_0 vision projector for vision capabilities.